基于D-S证据融合的飞机目标智能化虚拟仿真分类  

Intelligent Virtual Simulation Classification of Aircraft Targets Based on D-S Evidence Fusion

作　　者：陈玉洁 钟龙怀 黄龙亮 CHEN Yujie;ZHONG Longhuai;HUANG Longliang(Guizhou Equipment Manufacturing Polytechnic,Guiyang,Guizhou 550000,China)

机构地区：[1]贵州装备制造职业学院,贵州贵阳550000

出　　处：《自动化应用》2025年第6期84-87,共4页Automation Application

摘　　要：针对飞机目标的几何特征,研究了飞机目标的分类识别方法,提出了基于D-S证据融合的目标分类方法,再通过智能化虚拟仿真平台,对该目标方法进行仿真验证。首先,构建了各种证据(即各飞机目标的几何特性)的基本概率赋值函数;其次,输入待分类的飞机目标的各类几何特征值,分别计算出以每个特征值作为证据时,飞机目标属于各类别的概率;然后,利用一定的融合规则,将利用各类几何特征作为证据得到的飞机目标分类结果进行融合,并利用决策准则得到待分类的飞机目标所属类别;最后,利用智能化虚拟仿真平台,对所提目标方法的有效性进行仿真验证。Aiming at the geometric features of aircraft targets,a classification and recognition method for aircraft targets was studied,and a target classification method based on D-S evidence fusion was proposed.The target method was then simulated and verified through an intelligent virtual simulation platform.Firstly,basic probability assignment functions were constructed for various types of evidence(i.e.geometric characteristics of aircraft targets).Secondly,input various geometric feature values of the aircraft target to be classified,and calculate the probability that the aircraft target belongs to each category when each feature value is used as evidence.Then,using certain fusion rules,the aircraft target classification results obtained by using various geometric features as evidence are fused,and the decision criteria are used to obtain the category to which the aircraft target to be classified belongs.Finally,the effectiveness of the proposed target method is verified through simulation using an intelligent virtual simulation platform.

关 键 词：目标分类 识别 D-S证据融合 基本概率赋值函数 智能化虚拟仿真 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]